Claim Missing Document
Check
Articles

Found 2 Documents
Search
Journal : Jurnal Algoritma

Perbandingan Model SpaCy dan BERT untuk Persebaran Penggemar di Platform X (Twitter) Rahmadani, Nurul; Umam, Khothibul; Dwi Yuniarti, Wenty; Rini Handayani, Maya
Jurnal Algoritma Vol 22 No 2 (2025): Jurnal Algoritma
Publisher : Institut Teknologi Garut

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33364/algoritma/v.22-2.2310

Abstract

This study was conducted to compare the performance of the SpaCy Named Entity Recognition (NER) model and the Bidirectional Encoder Representation from Transformers (BERT) model in identifying the distribution of Bernadya fans based on the mention of Geo-Political Entity (GPE) locations. The dataset used was collected from X users' tweets using a scraping method with Python and will be analyzed on both NER models. The SpaCy NER model will be built from scratch with manual annotation, while the BERT model will be built using the transforms approach. From the evaluation results, the SpaCy model achieved a precision of 1.00, a recall of 0.92, and an F1-score of 0.96 on the training data, as well as a recall of 0.98 and an F1-score of 0.99 on the test data. The BERT model recorded a precision of 1.00, a recall of 0.95 (training), and 1.00 (testing), with an F1-score of 0.98 and 1.00. The Spacy model can recognize more than two entities well in one test sentence. However, when tested with the entire dataset, it cannot consistently recognize GPE entities. Conversely, the BERT model is better at recognizing GPE entities, with 4 GPE entities identified, including: Karanganyar, Indonesia, Mongolia, and Bandung as regions capable of identifying GPE entities with the most mentions. Therefore, in this study, the BERT model is better at recognizing GPE entities from the dataset used.
Pengelolaan Stok Lemari Kaca dengan Sistem SCM Berbasis Web untuk Efisiensi Produksi Apriliani; Rahmadani, Nurul; Suparmadi
Jurnal Algoritma Vol 22 No 2 (2025): Jurnal Algoritma
Publisher : Institut Teknologi Garut

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33364/algoritma/v.22-2.2495

Abstract

Utama Aluminium is a business engaged in the sale of glass cabinets. The problem faced in the operational activities of this business is irregularity in the management of raw material supplies. This condition is caused by inconsistency in supply from suppliers, which results in stock shortages and delays in orders, and on the other hand, there have also been instances of excess raw materials. The purpose of this study is to design a raw material inventory management application system for the glass cabinet production process, in order to optimize the management of raw material inventory, which has been a challenge in production activities at Utama Aluminium. The method used in this study is a descriptive qualitative method, with a case study approach at the company concerned. The results of implementing a web-based supply chain management (SCM) system showed a significant increase in efficiency: process time decreased by up to 40% through system automation and integration, operational costs were reduced by up to 25% due to labor efficiency and a decrease in error rates, stock accuracy increased to 95% with supplies matching demand, lead time was reduced by an average of 2 days, and the number of product returns decreased by up to 30% thanks to more accurate data information. This research can help the MSME sector improve distribution efficiency and inventory accuracy.